A Comprehensive Robust Bi-objective Model and a Memetic Solution Algorithm for Designing Reverse Supply Chain Network under Uncertainty

One of the major issues in a reverse supply chainmanagement is dealing with returned productsthrough a reverse logistics while considering uncertainty and non-financial performance measures alongside the more traditional financial measures. In this paper, a new comprehensive seven-layer recovery network is designed, including primary customers, collection/redistribution centers, recovery, recycling and disposal centers, and secondary customers. Uncertainties of quantity and quality of the returned products are considered. The aim of this paper is to design a reverse network such that both of total profit of the supply chain and responsiveness to customers’ demand are maximized during planning periods. Since the proposed mixed-integer linear problem belongs to the network design class of problems which is NP-hard, a memetic algorithm is developed that incorporates non-dominated sorting algorithm II and variable neighborhood search to find the Pareto-optimal solutions. The Taguchi experimental design technique is employed for parameter tuning of the proposed memetic algorithm. Efficiency of the proposed memetic algorithm is evaluated by comparing its performance with two other solution algorithms. Experimental results indicate the efficiency of the new memetic algorithm to solve the proposed bi-objective model for designing the reverse supply chain network under uncertainty.